A while back, I posted about my experience using BIAB RealTracks to accompany a solo instrument and vocalist, for upcoming live gigs, as well as for recording associated demo material. That thread can be found here:
http://www.pgmusic.com/ubbthreads/showflat.php?Cat=0&Number=225589&an=0&page=0You can hear samples of the music at
http://www.myspace.com/elegancemusicinfo Among the forum members who posted positive comments about my music was Peter Gannon, who also suggested that I post some details about the use of BIAB, and the recording session, to the Tips and Tricks forum. After some procrastination, here it is!
First, a disclaimer - I do not consider myself an "expert" in BIAB, or any of this stuff, for that matter. I'm still learning BIAB features and capabilities, like most of you. Yes, I've been doing audio for a long time, using various tools, but with the tools becoming as complicated as BIAB or modern DAW software, some concepts may have slipped through the cracks in my brain. So you may see some of what I'm doing as questionable or ridiculous - if so, please let me know so that I can stop my foolishness and do things in a better (or more logical) way. I'm always learning, and the PG Music forums have been, and continue to be, a great resource for my continuing education. Thanks!
Also, this is not an exhaustive diary of my setup & the recording session – it’s just an overall description to give any interested folks a fairly complete idea of how you could get results similar to mine, should you choose to tackle a project like this.
Even though BIAB is complicated and feature-laden, it's a testament to PG Music and the Band-In-A-Box developers that we can get excellent results in BIAB by knowing only some of the basics of the software.
Because there is a lot of information to post, I'll divide this up into four sections:
1. Hardware & software used
2. Session Prep
3. The Recording Session
4. Editing & Mixing
=========================
1. Hardware & software used:
Hardware:
• Laptop computer: emachines M6809, which is about six years old. A couple of years ago, I upgraded the RAM from the original 512 MB to 1.25 GB, and the internal hard drive to a UDMA, 150 GB 7200 rpm Hitachi. However, my working drive on this system is a 1 Terabyte external eSATA hard disk interfaced on a Cardbus eSATA card. A PCIe eSATA card would be a better choice, for those of you who have a current laptop with a PCI-express port, but like I mentioned, the M6809 is an old laptop. I think some current laptops also have an eSATA port built-in, which is even more convenient. I'm running Windows XP Pro SP3 on this machine.
• The M6809 also has a built-in firewire-400 port which has come in handy to connect the laptop to my audio interfaces - a Digidesign 002 control surface/audio interface, and my generic interface, an M-Audio.
• Audio Interface: M-Audio ProFire 2626. This is a capable firewire-based audio interface with plenty of I/O for one-man-band projects or other small groups. I won't do a commercial for it here, but it has worked well for me in various projects. No complaints.
• Microphones: I have a variety of mostly budget dynamic and condenser microphones, such as 2 Apex 435 condensers, a couple of Apex 135 small-diaphragm condensers, Oktava MK-012 small-diaphragm condensers, a couple of Shure SM-57 mics, etc.. But, getting to the point, the mic used on the flute for these sessions was an ElectroVoice ND767A hand-held super-cardioid dynamic, which is usually used as a hand-held vocal mic. But this mic has exceptional audio quality for a hand-held dynamic, so I felt comfortable using it to record with. I was pleased to discover that the mic Paula likes to use was also a high-quality ElectroVoice hand-held, an ND967 super-cardioid dynamic, similar to the flute microphone. So we used the ND967 to record her vocals.
• P.A. System: Carvin XP650-LM12. We'll use this on live gigs, but we also use it in our recording sessions for monitoring- no headphones! I'll get into that a little more in the section about the recording session. The XP650-LM12 is good basic P.A. for small to medium venues.
• Monitors (for the editing/mixing phase): Mackie HR 824 studio monitors. I also use an old pair of Sony headphones for additional monitoring, especially when I need to crank the level up to hear fine details in the sound without causing our two cats, and my wife, to run for the hills.
• Various mic cables, speaker cables, and mic stands. Not much, really- just a couple of mic stands for me and Paula, and my interconnections between pieces of equipment.
• Rectangular folding table, about 2' x 4' or thereabouts. I brought the equipment to Paula's house for the recording, so I needed something to put the stuff on.
Software:
• Band-In-A-Box (well, DUH!) The version I have is the 2009 UltraPak "Audiophile Edition", which is shipped on the external USB hard disk. Being picky about audio quality, I feel that the Audiophile Edition of BIAB is worth the extra $$, because you get full-quality .wav file versions of the RealTracks. This is a RealAdvantage (is that a PG Music copyright, Peter? :-) when you're creating your own works for creating CDs and other high-quality media. Even if you're using budget microphones for your live instruments or vocals, you might as well get the best quality sound you can on the backing tracks. Having said that, the EV mics we use are not slouches in the audio quality department. I can recommend them for all but the most critical recording applications. If it sounds good, it IS good. :-)
• Reaper. This is an excellent, low-cost DAW program (Digital Audio Workstation software) which works similarly to PG Music's PowerTracks Pro Audio, Logic, Pro Tools, Cubase, or any other modern DAW.. It was created and is still actively being developed by the same programmer (Justin Frankel) who created the WinAmp media player and the peer-to-peer file networking system called Gnutella. You might have noticed above that I have a Digidesign interface - the 002, and you may be wondering why I didn't use Pro Tools for this project. It's mainly because the current version of Pro Tools (8) is rather resource intensive, and my aging laptop is not quite up to the challenge. With Reaper, I can run more tracks and more effect plug-ins, without running out of CPU power. Plus, I love to try out new software, and while Reaper is not brand-new software, I hadn't used it before in a live recording session.
=========================
2. Session Prep
Before the session, we of course had previously discussed which songs we wanted to rehearse and record. For each song, within Band-In-A-Box, I did the following:
• Selected a RealTracks style that best seemed to fit the song, or my concept of the song, anyway
• Arranged the song - set the number of choruses, tempo, and key
• Generated an intro for most of the songs, and an ending. Well, most of the songs already came with endings, but I checked them.
• While I'm thinking about it, most of the songs we're using at this point I got from Wesley Dick's excellent BIAB page at
http://www.wesleydick.com/Songlist.html - the Real Book collection is super! But there are also other good song collections there.
• I usually mute all instruments in BIAB except for the RealTracks. We haven't used the RealTracks solo instrument much, since I've been soloing on flute, so I keep this BIAB instrument muted when we're recording. But I may use the solo RealTracks instrument live, occasionally.
• Hit the "Render to WAV file" button. In the "Render to Audio File" window, I check the "One WAV file per track" option. I also select the options "Include 2-bar lead-in" and "Include drum count-in sound", so that we can hear a count-off for the songs in the recording session.
The above was just about all the prep necessary for song prep, in our situation.
Before I imported the .wav files from BIAB into Reaper projects, I set up a template session in Reaper:
Track 1: Piano
Track 2: Drums
Track 3: Bass
Track 4: Vocal
Track 5: Flute
Track 6: Reverb
Track 7: Stereo Master Track
I then went into Reaper, and for each BIAB song, I created a new project file in a separate folder, imported the .wav files I had rendered in BIAB, and verified that I had the files in the project okay. In a couple of cases, for some reason, the files were imported in a different order, so that, for example, the bass .wav file got imported to the drums track, and the like. So for each Reaper project, I verified that the .wav files were on their correct tracks, moving them if necessary.
Before the session, I also defragged my system drive and the audio drive with a nice little (free!) defragger called jkdefrag. This seems to work better than the stock MS defragger in Windows XP, so I recommend it.
http://www.kessels.com/Jkdefrag/=========================
3. The Recording Session
On one of the sessions (we've done three so far), I set up one of the Apex 435 condensers as a second mic for Paula's vocal, running it to a separate track. I also set up a stereo pair of Octava MK-012 small-diaphragm condenser mics in an X-Y configuration. I positioned the X-Y mics about 10’ in the air in the middle of the room. Paula has a wonderful-sounding living room with a vaulted ceiling, and I wanted to try to capture some of this room sound on a separate stereo track.
As it turned out, I didn't use the room mic tracks much in the mixes- I felt that it didn't add much to the sound, so in most cases, I left these tracks muted. But this kind of thing is always worth a try, if you have the extra mics.
The mics were all plugged into channels of my M-Audio ProFire 2626 interface, which is connected to the laptop via firewire. Before we actually recorded, I got input levels for me and Paula. I like to set mic levels by having the voice or instrument play as loud as they're going to sing (or play), seeing where the peaks are on the clipping indicator, and then backing off on the input level control about 10 db, to account for peaks and exuberance in performance. In a modern digital system like this, you can probably run the inputs -15 or -20 db below clipping, which would give you plenty of headroom for peaks and still have a good S/N ratio. You want to make sure you don't have the signals hot enough that they clip on peaks- digital clipping sounds nasty!
For monitoring, I had the stereo output from Reaper coming out of channels 1 and 2 on the ProFire 2626 and into two channels of my P.A. system, a small Carvin P.A. with 12" speakers. I set the volume on these low, but enough for us to clearly hear the rhythm tracks when we recorded. That's right- we didn't use any headphones in this session for monitoring- only speakers! The reason I could get away with this is because:
a) both of my sound sources were being close-miked, (with the exception of the room mics) and
b) the P.A. speakers were turned down to a minimal level, so that not much of the P.A. was leaking into the recording mics.
c) I also positioned the speakers so that the mics were pointing away from the speakers, to cut the level from them further. Both mics we used are supercardioid, so they have a good amount of rejection of sound that is coming to them from the back.
The upshot of using the live P.A. for monitoring is that if, for example, I listen to the flute track when there is no flute on the track, I can her the P.A. (and Paula, if she's singing) at a very low level on the track; not enough to affect the mix, though. I would say it’s more than 40 db down from the level of the flute when it’s playing. This is great, because I'd rather not use headphones when I'm recording. And I'll bet a lot of other performers feel similarly. The main thing to keep in mind if you do this, is that you have to keep the level of the P.A. down to a minimum- just enough to monitor what’s going on, and no more. And use close-mic technique, with directional mics, and keep the mics pointed away from the speakers.
Reaper is quite flexible with track configuration and signal routing - in this session, I used sends from the piano, vocal, and flute tracks, to a reverb track.. The reverb track had one effect set up in it (reverb, obviously!) with the effect mix set to 100% "wet". This way, I can use the fader on the reverb track to control the overall amount of reverb, and control how much of each track gets reverb with the send controls within the tracks’ i/o section.
It's probably easier to show this configuration than to try to explain it! If you use Reaper for a week or two, you will see how easy it is to set up various routings to and from tracks. The graphical interface is quite different from Pro Tools, but most of the concepts and functionality are the same. If you learn one DAW, it’s fairly easy to pick up another one, for this reason. As it is to pick up additional musical instruments after you’ve learned one.
Most songs had three instrument tracks imported from BIAB - piano, acoustic bass, and drums. One or two songs had guitar instead of the piano, and a couple of them had both piano and guitar tracks (so there's a four-piece rhythm section in these cases).
Paula and I used the EV mics as we would in a live situation- a couple of inches to maybe 6 inches from the mics. By the way, I put foam windscreens on both mics, to minimize any "P-popping” or other breath noises. A couple of P-plosives still snuck through on Paula’s vocal, but it’s not very noticeable. About the only place I noticed it was when she sings “pitter-patter” in “When Sunny Gets Blue”. I could cut some of these plosive sounds after-the-fact in the mix, by “scooping out” the thump frequencies on just those plosives, but I didn’t bother in this case.
Of course, the RealTracks in BIAB are sampled, professionally-recorded tracks, and they seem to be of high quality. Off the top of my head, I can't tell you the bit rate of the Real Tracks, but I believe they are 24-bit, 48-kHz samples. I didn't use any EQ on the BIAB-imported tracks; just a bit of high-frequency boost on the master stereo track (about 3 db at 10 kHz). I think that the challenge is to come up with tracks for your live instruments or vocals which match the high quality of the BIAB RealTracks. With budget mics, that’s even more of a challenge.
The actual recording of these songs was straightforward- I armed the tracks for recording in Reaper, and started the songs in record mode. We only did retakes in a couple of cases. Most of the songs were recorded straight through from beginning to end, as we would play them on a gig. There were a few mistakes and glitches in our performance, of course, which I attempted to fix in the Editing & Mixing phase.
=========================
4. Editing & Mixing
I have a particular way, or workflow, in which I usually do mixing. I’m sure everybody develops their own system. Whatever works best for you is probably the “right” way of doing things. I’ve seen some good thoughts about this elsewhere on this forum.
First, for most of the BIAB songs, I used the arrangement as output by BIAB in Reaper. There was one major exception to this- when we recorded "When Sunny Gets Blue", we realized that the song was too long, when we recorded it- three times through the song form. When I mixed this song, I did something slightly unorthodox for this type of music- I changed the song form! Instead of going into a third time through the song, I cut to the final bridge at that point with the flute, and ended the song after the verse part. So it's really two and a half times through the song form. It sounds slightly jarring to go into the bridge when you expect another verse, but I've gotten used to it that way.
Before I mix, I usually "normalize" all audio tracks, so that they are similar in peak levels. Yes, I know that the *average* levels of the tracks may vary greatly, depending on the source material, and this practice doesn’t improve the sound in any way, but it’s just something I’ve gotten in the habit of doing. For tracks with similar content, I feel it helps me get a more or less consistent starting point for setting the levels.
Then, I play only the rhythm section tracks, getting a good relative balance between them (in loudness). I sometimes do a bit of mixing, then I’ll go listen to something else, like a Steely Dan CD or some other material which has been expertly mixed and mastered, to get a perspective on my own mix when I start back in with it.
After I’ve gotten a relatively good balance among the rhythm section instruments, I add solo instruments or vocals to the mix and try to get everything to play nice together. It's not always easy, but in this session, it seemed to come together very easily. I'm by no means saying I'm a world-class recording or mixing engineer, but I think I got a fairly pleasing balance on these mixes. A lot depends on what you’re monitoring the mix on, too- I switch between headphones and my Mackie monitors quite often. I really should add a small radio speaker as a third monitor, but I haven’t incorporated that into my mix of equipment (yet). But I do throw mixes onto a CD and listen in my car, before finalizing them, which gives me a different sonic perspective on the mixes. My main point is, your mix may sound great on expensive studio monitors, but that kind of equipment is not what most people will listen to your material on. Hence the importance of monitoring on everyday, “less than optimal” equipment, to make judgments on spectral balance, in addition to your studio monitors.
On these sessions I did some peak limiting, and about 6 db of compression, on the vocal, flute, and Master tracks. I used the VST “Reacomp” plug-in for the comp/limiting, and the “ReaEQ” for the EQ. These plug-ins come with Reaper. On some mixes, I used the DX limiter that came with BIAB, which works pretty well. I also used about 3db of boost at 200 Hz and 4 db of boost at 10 kHz on the vocal track, to give it some warmth and sparkle. I've found that a minimalist approach on the signal processing usually gives the most "natural" sound, assuming you're using good mics (or high-quality sampled instruments) and are careful with keeping the signal gain optimal for each channel source. I believe that on the vocals, I might have gotten a better "sheen", or "silky" high-end, on her voice, if I had used a top-quality studio condenser mic, but firstly, I don't have one, and like I said, I (and she) was pleased with the sound of her EV ND967, so I didn't consider miking her differently.
I did some editing to correct timing errors- not too much of this, though. In a few cases, I also recorded additional flute takes and used some of those instead of the take recorded at the original sessions.
I also recently got a plug-in which I feel is quite handy for fixing small pitch discrepancies and timing errors- Celemony’s “Melodyne” plug-in. Similar to Antares’ famous “Auto-Tune” software, Melodyne has most of the same functionality, and I really like the graphical interface. If you’re not familiar with it, go to the page at
http://www.celemony.com/cms/index.php?id=plugin1&L=0 and play the four videos to get an idea of its functionality and interface. As with any pitch-manipulation software, there is a temptation to overuse it, to make things too perfect-sounding. I try to keep my usage of this plug-in to a minimum, only correcting really noticeable pitch problems. Thankfully, Paula is a good vocalist, and I didn’t have to use much of it. Just nudged a few notes here and there. I actually changed a couple of flute notes to different pitches, to reflect what I had intended to play, but my fingers got in the way. ;-) I challenge anyone to listen to the vocals on the songs and see if you can detect any of the pitch manipulation- if you use it correctly, Melodyne can be very subtle in its processing of the audio you run through it. If any of my pitch processing can be detected, chalk it up to my carelessness with the controls, not the plug-in itself.
The reverb plug-in I used was the DX-based plug that comes with BIAB. I wouldn’t call this a top-quality reverb, but it is adequate for most uses. I’ll probably want to get a high-quality VST reverb plug-in for critical use, such as producing a CD. I was driven more by time constraints than quality considerations in these mixes, but rest assured that I will take more time and care in the entire process, when I produce a CD.
If you'll recall from my statement above, I had rendered the .wav files from BIAB with the two-bar drum count-in, so that we could hear the count-in while adding the vocal and flute tracks. However, I didn't want to leave the count-in in the final mix, so I edited the five audio tracks by removing all the audio at the beginning of the tracks, up to just after the beat 4 click on the bar before the music starts. I was surprised to still hear the echo of the beat 4 click after I had removed it, and found that this audio was coming from the drum track. So I separated the small section of the drum track just before the drum entrance, and muted that section of the track, so that there would be complete silence for the fraction of a second before the music starts.
In sections of the song in which the flute was soloing and there was no vocal, I generally muted the vocal track. This was especially needed on the songs in which I used a different flute take than the original. Where there was a different flute part than the one on the original recording, the original flute part could be heard on the vocal track- since we were recording the vocals and flute in the same space, there was a lot of leakage of the vocal into the flute mic, and vice-versa. So in these cases I muted the vocal track during the flute solos. In some cases, I left the vocal track unmuted during the flute solo (when I was using the original flute take, because in those cases, it didn't matter that a little of the flute was leaking onto the vocal track).
For rendering to .mp3 or other file format outputs, Reaper has the 'Render' choice from the main menu. This has plenty of quality options, and I outputted the mixes to .wav files first, and then to .mp3, since I was going to be uploading them to MySpace. In the .mp3 renderings, I used "high quality" options, and if I remember correctly, I used a bit rate of 192 kbps or higher for these files. On the longer songs, I had to drop down to 160 kbps to keep the resulting file sizes below 6 MB, since MySpace has a size limit of 6 MB for the files you upload to the player. The .mp3s came out pretty clean-sounding, as you can hear.
----------------------
That’s about it for my description of the recording sessions and how I used BIAB RealTracks in the production. It is not difficult, and I think that most BIAB users should be able to get similar results in their own projects. I hope I didn’t leave out any important details, but let me know if I did, and I’ll fill in the omissions. Also let me know if you have any further questions- I’ll be glad to answer them.
Regards,
Doug
Dansk
Deutsch - Windows
Español
Français - Windows
Italiano - Windows
Nederlands
Polski
Português
Svenskt
русский
Japanese
Simplified Chinese
Traditional Chinese
Korean